In the previous lesson, you looked at your work and identified places where AI could help.
Now we’re going to take the next step.
Instead of thinking about one AI task at a time, start thinking about the whole process.
That’s where the idea of an AI workflow becomes useful.
What Is a Workflow?
A workflow is simply a series of steps that lead to a result.
You already use workflows every day without necessarily calling them that.
For example, making a cup of coffee might be:
Get the cup → make the coffee → add milk → drink it
Work can be more complicated.
Preparing a customer report might be:
Collect information → organise it → identify important points → write the report → check it → send it
That’s a workflow.
Where Does AI Fit?
Now look at that process again.
Perhaps AI could help organise the information.
Perhaps it could identify the important points.
Perhaps it could create a first draft.
But you might still want to check everything before the report is sent.
So the workflow could become:
Information → AI organises → AI helps identify key points → AI creates draft → You review → Final report
Notice what has happened.
You haven’t simply asked AI:
“Write my report.”
You’ve thought about how AI fits into the process.
That gives you more control over the result.
One Step Can Feed the Next
A useful workflow is connected.
The result from one step becomes useful information for the next step.
For example:
Meeting notes
↓
AI organises the notes
↓
Key decisions and actions
↓
AI creates an action list
↓
You check the list
↓
Final version
Each step has a purpose.
You aren’t asking AI to do random things.
You’re building a path toward a result.
You Still Decide Where You Belong
An AI workflow doesn’t mean AI does everything.
In fact, some of the most useful workflows combine:
AI work
Your work
AI + your judgement
For example:
AI: Organise the information.
You: Decide what matters.
AI: Create a draft.
You: Check and improve it.
This can allow AI to handle some of the time-consuming work while you focus on the parts that need your experience and judgement.
Don’t Add AI Just Because You Can
This is important.
If a step already works well and AI wouldn’t improve it, leave it alone.
You don’t need to make every step “AI-powered.”
A good workflow is not the one containing the most AI.
It’s the one that produces the best useful result with the least unnecessary effort.
Start Small
You don’t need to redesign an entire business process.
Start with something manageable.
For example:
Preparing a weekly team update
or
Turning meeting notes into an action list
or
Researching a topic before a meeting
Once you’ve created one useful workflow, you can improve it over time.
Let’s Make It Practical
Go back to the task you mapped in Lesson 1.
Look at the steps again.
Now ask:
Which steps could connect together with AI helping at one or more points?
Write the process as a simple chain.
For example:
Collect → Organise → Analyse → Draft → Review → Share
Then mark where AI could help.
Finally, ask:
“What do I want the final result to look like?”
That last question is important.
A workflow isn’t just a collection of AI instructions.
It is a path toward a useful result.
What You’ve Learned
You’ve moved from:
“AI could help with this task.”
to:
“I can see how AI could fit into the whole process.”
That’s the beginning of workflow thinking.
You now know how to:
- break a task into steps
- identify where AI can help
- connect useful AI-assisted steps
- keep human judgement where it matters
- avoid adding AI where it doesn’t add value
But there’s an obvious next question.
If you’ve created a workflow that works well, why rebuild it from scratch every time you do the same job?
That’s where we go next.
Next Lesson